U.S. patent application number 14/356630 was filed with the patent office on 2014-10-09 for method and apparatus for creating cost data for use in generating a route across an electronic map.
The applicant listed for this patent is TOMTOM NORTH AMERICA, INC.. Invention is credited to Clayton Richard Morlock.
Application Number | 20140303892 14/356630 |
Document ID | / |
Family ID | 46052043 |
Filed Date | 2014-10-09 |
United States Patent
Application |
20140303892 |
Kind Code |
A1 |
Morlock; Clayton Richard |
October 9, 2014 |
METHOD AND APPARATUS FOR CREATING COST DATA FOR USE IN GENERATING A
ROUTE ACROSS AN ELECTRONIC MAP
Abstract
A method is disclosed involving receiving GPS data from persona
portable training devices of users when traversing an off-road
segment of an electronic map together with associated data
indicative of a heart rate of a user during the movements. The
position and heart rate data for each user traversing the segment
are processed using data indicative of a fitness profile for the
user. The resulting data is used to determine a normalized cost to
be associated with the segment, indicative of the difficulty in
traversing the segment. The cost data is generated using a neural
network. The resulting cost data for different segments in a
network of segments is used to generate route suggestions for users
based upon desired workout intensity, fitness levels, etc.
Inventors: |
Morlock; Clayton Richard;
(Lebanon, NH) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
TOMTOM NORTH AMERICA, INC. |
Lebanon |
NH |
US |
|
|
Family ID: |
46052043 |
Appl. No.: |
14/356630 |
Filed: |
November 19, 2012 |
PCT Filed: |
November 19, 2012 |
PCT NO: |
PCT/US12/65742 |
371 Date: |
May 7, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61561345 |
Nov 18, 2011 |
|
|
|
Current U.S.
Class: |
701/533 |
Current CPC
Class: |
G01C 21/20 20130101;
G06T 17/00 20130101; G06T 17/05 20130101; A63B 2220/62 20130101;
A63B 69/0028 20130101; A63B 2230/06 20130101; G06F 16/29 20190101;
A63B 2220/30 20130101; G09B 29/007 20130101 |
Class at
Publication: |
701/533 |
International
Class: |
A63B 69/00 20060101
A63B069/00; G09B 29/00 20060101 G09B029/00; G01C 21/20 20060101
G01C021/20 |
Claims
1. A method for creating cost data for use in generating routes
across an electronic map of navigable segments, the method
comprising: receiving probe data from a plurality of users, the
probe data comprising, for each user, movement data indicative of
the movements of the user with respect to time, and physical
exertion data indicative of at least one measure of physical
exertion associated with the movements of the user over time;
processing the received probe data for each user using an ability
profile for the user to generate normalized routing cost data for
one or more of the navigable segments of the electronic map; and
associating the routing cost data with the one or more navigable
segments of the electronic map.
2. The method of claim 1, wherein the probe data comprises movement
data indicative of the movements of the multiple users along one or
more given navigable segments of the electronic map and associated
physical exertion data, and wherein the method comprises processing
the probe data from the multiple users relating to travel of the
users along the one or more given navigable segments using the
ability profile for each user to enable normalized routing costs
for each of the one or more given navigable segments to be
generated.
3. The method of claim 1, wherein the movement data comprises data
indicative of the position of the user with respect to time.
4. The method of claim 3, wherein the position data is received
from a location determining and tracking means of a personal
portable training device arranged to be transported, carried or
worn by a user, and wherein the location determining and tracking
means comprises a global navigation satellite systems (GNSS)
receiver.
5. The method of claim 1, wherein the physical exertion data is
indicative of a heart rate of the user.
6. The method of claim 1, wherein the ability profile is a function
of at least a fitness level of the user.
7. The method of claim 1, wherein the cost data is determined using
a machine learning process.
8. The method of claim 1, wherein the cost data for the one or more
navigable segments is at least one of time and weather
dependent.
9. The method of claim 1, wherein the or each navigable segment is
an off-road segment.
10. The method of claim 1, further comprising using the normalized
cost data to provide at least one of: a suggested route to a user
based on an ability profile of the user; and one or more user
specified parameters.
11. The method of claim 10, wherein the one or more user specified
parameters are selected from the group consisting of: activity
time, distance, start position, end position, segment type, and
level of physical exertion.
12. A non-transitory computer readable medium including a computer
program comprising computer program code means adapted to perform
the method of claim 1 when said program is run on a computer.
13. (canceled)
14. An electronic map comprising data representative of a plurality
of navigable segments, and routing cost data associated with one or
more of the navigable segments created using the method of claim
1.
15. A system for creating cost data for use in generating routes
across an electronic map of navigable segments, the system
comprising one or more processors, said one or more processors
being arranged to: receive probe data from a plurality of users,
the probe data comprising, for each user, movement data indicative
of the movements of the user with respect to time, and physical
exertion data indicative of at least one measure of physical
exertion associated with the movements of the user over time;
process the received probe data for each user using an ability
profile for the user to generate normalized routing cost data for
one or more navigable segments of the electronic map; and associate
the routing cost data with the one or more navigable segments of
the electronic map.
16. The method of claim 3, wherein the movement data further
comprises elevation data indicative of the elevation of the user
with respect to time.
Description
FIELD OF THE INVENTION
[0001] This invention relates to methods and systems for
determining cost data for use in generating a route across an
electronic map. The method is particularly, although not
exclusively, concerned with determining cost data in respect of
traversing navigable segments of an electronic map that are
off-road segments. The invention provides methods and systems for
determining cost data associated with the traversal of segments of
the electronic map at least partially under a user's own power.
BACKGROUND OF THE INVENTION
[0002] Portable navigation devices (PNDs) that include GNSS (Global
Navigation Satellite Systems) signal reception and processing
functionality are well known and are widely employed as in-car or
other vehicle navigation systems. Such devices include a GNSS
antenna, such as a GPS antenna, by means of which
satellite-broadcast signals, including location data, can be
received and subsequently processed to determine a current location
of the device. The PND device may also include electronic
gyroscopes and accelerometers which produce signals that can be
processed to determine the current angular and linear acceleration,
and in turn, and in conjunction with location information derived
from the GPS signal, velocity and relative displacement of the
device and this vehicle in which it is typically mounted. Such
sensors are most commonly provided in in-vehicle navigation
systems, but may also be provided in the PND device itself.
[0003] In recent years, GPS has started to be used for pedestrian
and outdoor applications. Currently there are a significant number
of portable personal training devices having location determining
capabilities available on the market for use e.g. in sporting
activities, pedestrian travel and other outdoor applications.
Portable personal training devices are devices that may be worn by
runners, joggers, cyclists and other athletes and outdoor
enthusiasts, etc, and which can track and record data indicative of
the movement of the user over time, e.g. the pace of the user at
particular moments during a workout and/or the distance covered by
the user during the workout. For example, sports watches that
include GPS antennas have started to be used by such users as a
means to obtain real-time data of their speed, distance travelled,
etc. The GPS data is also typically stored on such devices such
that it can be analysed after the athlete has finished their
activity, e.g. in some cases by transferring the collected data to
a computer or website to be displayed as traces on a digital map
(i.e. in a form of Geographic Information System (GIS)).
[0004] Such portable personal training devices can also be linked
with sensors, such as a heart rate sensor (that is typically worn
on a strap around a user's chest), such that heart rate data can be
collected and stored on the device during a period of exercise. The
collected data can then be transferred from the device to a
computer or website, together with the GPS data, to allow a user to
see their level of physical exertion whilst undertaking the
exercise.
[0005] It is therefore known in the art to track a course for a
workout, e.g. using a portable personal training device, and
compare user physical parameters such as heart rate, travel time
and other parameters between individuals and between separate trips
for a single individual.
[0006] However, the Applicant has realised that there remains a
need for a method and system that will allow routes to be proposed
to users for navigation by the user under their own power, and, in
particular, which involve off-road navigable segments. There is
also a need for improved methods of generating an electronic map
comprising off-road navigable segments.
SUMMARY OF THE INVENTION
[0007] In accordance with a first aspect of the invention there is
provided a method for creating cost data for use in generating
routes across an electronic map of navigable segments, the method
comprising:
[0008] receiving probe data from a plurality of users, the probe
data comprising, for each user, movement data indicative of the
movements of the user with respect to time, and physical exertion
data indicative of at least one measure of physical exertion
associated with the movements of the user over time; and
[0009] processing the received probe data for each user using an
ability profile for the user to enable normalized routing cost data
for navigable segments of the electronic map to be generated.
[0010] In accordance with the invention in any of its aspects or
embodiments, the probe data comprises data indicative of the
movements of multiple users over time, and associated physical
exertion data relating to the movements. The movement of the user
is a movement carried out under the user's own power. In other
words, the movement is accompanied by physical exertion on the part
of the user. The movement is carried out at least partially, and
preferably entirely, under the users own power. The movement may be
any human powered movement. The movement may or may not be
associated with the movement of a user powered vehicle. In
embodiments involving a user powered vehicle, the user powered
vehicle may be a vehicle that is entirely or partially user
powered. The movement may therefore be the movement of a pedestrian
user, and may be a walking, jogging or running movement of the
user. In other embodiments the movement may be a movement of the
user when propelling a user powered vehicle such as a boat (e.g.
canoe, kayak, etc), bicycle, skis, etc. The vehicle may therefore
be a land or water based vehicle. The movement may be a movement
involved in a sporting, commuting or leisure activity.
[0011] In accordance with the invention, probe data is obtained
from each of a plurality of users which comprises a movement
component, indicative of the movements of each user with respect to
time, and a physical exertion component, which is indicative of the
level of physical exertion of the user when performing the
movements. Thus the probe data comprises movement data for each of
the plurality of users and physical exertion data associated
therewith. The probe data for each user is used together with an
ability profile for the respective user in the generation of
normalized routing cost data for navigable segments of an
electronic map.
[0012] The present invention thus provides a method of obtaining
normalized cost data that may be used for generating routes across
the electronic map based upon probe data obtained from multiple
users. The physical exertion probe data for a given user is
indicative of the level of difficulty for that user in traversing a
given navigable segment. It will be appreciated that as it is
carried out at least partially under the power of a user, the
traversal of the same navigable segment may present differing
levels of difficulty to different users depending, e.g. upon their
personal fitness level, experience level, etc. The present
invention takes into account such individual variation to enable
normalized, i.e. standardised, cost data to be generated for
segments by processing the probe data for each user together with
ability profile data for the user. The ability profile data
provides a way of scaling the probe data, e.g. physical exertion
data, to provide data for use in generating cost data for a segment
that is comparable for different users, and which enables
normalized cost data to be determined using the probe data obtained
from the multiple users. In this way, the navigable segments of the
electronic map may be globally graded with a cost indicative of the
relative difficulty of traversing the segment.
[0013] The present invention also extends to a system, optionally a
server, for creating cost data for use in generating routes across
an electronic map of navigable segments, the system comprising:
[0014] means for receiving probe data from a plurality of users,
the probe data comprising, for each user, movement data indicative
of the movements of the user with respect to time, and physical
exertion data indicative of at least one measure of physical
exertion associated with the movements of the user over time;
and
[0015] means for processing the received probe data for each user
using an ability profile for the user to enable normalized routing
cost data for navigable segments of the electronic map to be
generated.
[0016] The present invention in this further aspect may include any
or all of the features described in relation to the first aspect of
the invention, and vice versa, to the extent that they are not
mutually inconsistent. Thus, if not explicitly stated herein, the
system of the present invention may comprise means for carrying out
any of the steps of the method described.
[0017] The means for carrying out any of the steps of the method
may comprise a set of one or more processors for so doing. A given
step may be carried out using the same or a different set of
processors to any other step. Any given step may be carried out
using a combination of sets of processors.
[0018] The probe data may be used in accordance with the invention
to generate normalized routing costs for navigable segments of an
electronic map. The probe data received from multiple users is
indicative of the movements of the users over time, and their
associated exertion. In embodiments the data is indicative of the
movements of the multiple users along one or more given navigable
segments of the electronic map. The method may then comprise
processing the probe data from the multiple users relating to
travel of the users along the one or more given navigable segments
using the ability profile for each user to enable normalized
routing costs for each of the one or more given navigable segments
to be generated.
[0019] As will be discussed in more detail below, other data in
addition to the ability profile may be used to process the received
probe data so as to generate the normalized routing costs,
including: data indicative of historic weather conditions (e.g.
during the past few days or weeks); data indicative of current
and/or expected weather conditions; data indicative of the terrain
or general construction of the navigable segments being traversed
(e.g. concrete, gravel, mud, etc).
[0020] The one or more navigable segments referred to herein may be
of any type. Preferably the or each segment is an off-road segment.
A navigable segment may be any path that may be taken by a user
when moving at least partially under their own power, with or
without the associated movement of a human powered vehicle. A
navigable segment may be a water or land based segment. A navigable
segment may be a segment of a footpath, river, canal, cycle path,
tow path, railway line, or the like. The segment may be a segment
that is for use by a non-motorized vehicle such as a bicycle, or
skis, or by motorized vehicles such as an ATV
(all-terrain-vehicle), snowmobile, or off-road motorcycle. A water
segment may be a segment that is for use by a canoe, kayak or the
like. An off-road navigable segment may or may not be man made. For
example, such a navigable segment may follow a natural path
existing in a landscape and/or a path that has been made by other
users, e.g. a runner-made path through a forest.
[0021] The method comprises receiving probe data from multiple
users comprising movement data indicative of the movements of the
users along a given path comprising one or more navigable segments
with respect to time, and physical exertion data indicative of at
least one measure of physical exertion associated with the
movements along the given path over time. As will be appreciated,
the probe data to be processed may be received directly from, for
example, portable personal training devices of users. However, it
is also envisaged that the probe data may be obtained from a data
store. In other words, data transmitted from users may be stored at
least temporarily at a first server, where it may also be
pre-processed (e.g. position data may be smoothed, etc), and then
the stored, optionally pre-processed, probe data may be retrieved
for subsequent processing to create the normalised routing
costs.
[0022] In accordance with the invention position data and
associated exertion data are obtained for multiple users traversing
the same navigable segment. In embodiments the navigable segment
forms part of a trail that is followed by a user. Exertion data
relating to different users may be obtained in respect of the
traversal of a navigable segment as part of the same or different
trails followed by different users. A path as referred to herein is
made up of one or more navigable segments, which may be of any of
the types described herein. The term "trail" as used herein refers
to a path comprising one or more navigable off-road segments. The
one or more off-road navigable segments may be of any of the types
described above. Thus a trail may refer to an off-road trail, e.g.
a pedestrian path, recreational path, river path, etc. While the
present invention is described by particular reference to trails
i.e. off-road paths, and off-road navigable segments, it will be
appreciated that the methods of the invention may be applied to
navigable segments, and paths comprising navigable segments, that
are or include road type segments, e.g. permanent footpaths and the
like. If not explicitly stated herein, embodiments described by
reference to a "trail" may be applicable to any form of path made
up of navigable segment(s), unless the context demands
otherwise.
[0023] It will be appreciated that in contrast to road type
navigable segments, off-road navigable segments or paths, i.e.
trails, made up of such segments, may change over time, potentially
with some frequency. For example, new trails may be added, trails
re-routed, trails may degrade and/or be improved, etc. As the
present invention determines cost data for navigable segments, and,
in embodiments, the course of the navigable segments, using probe
data collected from actual users, the method may provide a dynamic
system that can provide accurate cost data in relation to an ever
changing network of segments, and subsequently may be used to
provide routing in such a changing network.
[0024] As described above, the present invention relates to
creating cost data for use in generating a route across an
electronic map of navigable segments. The navigable segments may be
known navigable segments. In other words, the navigable segments
may be predefined navigable segments of an existing electronic map.
In other embodiments, the method may extend to the step of
determining the navigable segments. The step is preferably carried
out using the movement component of the probe data, i.e. positional
data relating to the movements of users over time. The step may be
carried out before determining the cost data for the segments in
accordance with the invention. The method may therefore comprise
the step of generating the electronic map. Of course, a combination
of such techniques may be used, i.e. the electronic map may
comprise a known portion and a portion that is generated using the
probe data. In other words, the navigable segments of a
pre-existing electronic map are continuously being updated using
the received positional data from a plurality of users. In
embodiments therefore the electronic map comprises a known network
of navigable segments and/or the method comprises using the probe
data to generate a network of navigable segments providing at least
a portion of the electronic map.
[0025] In embodiments in which the method comprises a step of
generating at least a portion of the electronic map using the probe
data, the method may comprise using the movement data indicative of
the movements of each of the plurality of users with respect to
time in a geographic area to determine a network of one or more
navigable segments in the area. The method may comprise using the
movement data to determine one or more paths taken by a plurality
of users in the geographic area, and generating a network of one or
more navigable segments in the area. It will be appreciated that
positional traces, i.e. position against time, for the plurality of
users determined based on the probe data may be used to infer the
likely positions of paths and hence navigable segments in the area.
The method may involve clustering probe traces to determine
commonly taken paths, etc.
[0026] In accordance with the invention in any of its aspects or
embodiments, the received probe data may be received in any
suitable manner. The data may be received via any suitable
communications link. The link may be a wireless link or a wired
link or may comprise combinations thereof. For example, the data
may be received via the Internet or over the air. As discussed
above, the data may be received directly from personal portable
training devices of the users, or indirectly from such devices by
retrieving previously uploaded data from a data store.
[0027] The received probe data may be received from any suitable
source or sources. The received probe data comprises movement data
for each user with respect to time and associated physical exertion
data. The probe data for each user is received from a mobile device
or devices associated with a user. The device or devices may be
transported, worn or carried by the user. Thus, the movement of the
device or devices may be assumed to correspond to the movement of
the user.
[0028] In preferred embodiments the movement probe data is received
from the location determining and tracking means of a mobile device
associated with the user. The location determining and tracking
means may be arranged to determine and track the location of the
device. The location determining and tracking means could be of any
type. For example, latitude and longitude coordinates could be
determined using a device that can access and receive information
from WiFi access points or cellular communication networks.
Preferably, however, the location determining and tracking means
comprises a global navigation satellite systems (GNSS) receiver,
such as a GPS receiver, for receiving satellite signals indicating
the position of the receiver (and thus user) at a particular point
in time, and which receives updated position information at regular
intervals. In preferred embodiments the location determining and
tracking means comprises a global navigation satellite systems
(GNSS) receiver, preferably a GPS receiver, and preferably a GPS
chipset. Thus, in these particularly preferred embodiments, the
method comprises receiving the position data from a GPS chipset of
a device. In embodiments, therefore, position data is received from
location determining and tracking means of a personal portable
training device arranged to be transported, carried or worn by a
user, and wherein the location determining and tracking means
comprises a global navigation satellite systems (GNSS)
receiver.
[0029] The movement data relates to the position of the user at
different times, e.g. when travelling along a path. Thus the
movement data includes position data for the user at different
times. Accordingly the movement data may consist of a set of
position data points, each data point representing the position of
the user at a given time. Preferably the data comprises GPS data.
In embodiments the position data may be obtained by the device at
any given frequency to allow tracking of the device and hence user.
In some embodiments the position data is obtained by the device at
a rate of 0.5 Hz or greater, preferably at a rate of 1 Hz or
greater, such as up to a rate of 20 Hz. In some embodiments the
position data is received by the device at a rate of 1 Hz.
Accordingly, in embodiments, the position data may relate to the
position of a user at time intervals of less than or equal to 2
seconds, or less than or equal to 1 second, such as up to 0.05 s.
In embodiments the received movement data comprises position data
is associated with time information i.e. identifying the time to
which the position data relates. The time information may be in the
form of a timestamp.
[0030] The movement data is indicative of the movements of the user
with respect to time. The movement data is indicative at least of
the position of the user with respect to time. The position may be
a two dimensional position. However, in preferred embodiments the
movement data further comprises elevation data indicative of the
elevation of the user with respect to time. Thus the position is
preferably a three dimensional position. Thus the movement data
preferably includes at least longitudinal and latitude data, and
may include elevation data. Elevation data may be obtained in a
similar manner to longitudinal and latitude position data from a
GPS chipset of a device, or from a separate sensor, such as a
barometric sensor of the device. The steps of the method may be
carried out using two or three dimensional position data as
desired.
[0031] The device from which the movement data is received may be
of any suitable type. In preferred embodiments the device is a
mobile device that is arranged to be transported, carried or worn
by a user. Preferably the mobile device does not include navigation
functionality as found in vehicle PNDs. For example, preferably the
device does not include map data stored within a memory of the
device or processing means that can use map data to determine a
route between a first location (or "origin") and a second location
(or "destination") and provide suitable navigation (or guidance)
instructions.
[0032] In some preferred embodiments, the mobile device is arranged
to be carried by a user as he or she travels from one location to
another. The mobile device can be arranged so as to be carried by
the user, such as being attached to the user's arm or wrist, or
simply by being placed in a pocket or other suitable receptacle
(e.g. a specially designed holder or case). In other embodiments,
the mobile device can be arranged so as to be transported by a
user. For example, the mobile device can be attached to a vehicle
being used by the user, e.g. a bicycle, canoe, kayak or other
similar vehicle. The mobile device could also be attached to an
object being pushed or pulled by a user, such as a child-carrying
buggy. Such mobile devices are commonly referred to as portable
personal training devices. Thus, in particularly preferred
embodiments, the mobile device is a portable personal training
device. In some preferred embodiments, the mobile device is a
sports watch. Exemplary mobile devices from which data may be
received in accordance with the present invention are described in
the International (PCT) application no. PCT/EP2011/054686, filed on
28 Mar. 2011, and published as WO 2012/045483; the entire contents
of which is incorporated herein by reference.
[0033] The movement data received relates to the travel of a user
along a path. It will be appreciated the term "path" herein refers
to any journey or movement made by a user along one or more
navigable segments, and does not imply that the user has followed a
pre-planned route.
[0034] The user probe data additionally comprises physical exertion
data. The physical exertion data is indicative of one or more
measures of the physical exertion of the user and is associated
with the movement data. The physical exertion data is preferably
obtained from a sensor or sensors associated with a mobile device
from which the movement data is obtained, e.g. from a portable
personal training device. The sensors may be one or more "external"
sensors operably connected to the mobile device and which are
located outside a main housing of the device, or may be located
within a main housing of the device. In other embodiments, the
physical exertion data may be obtained from a separate mobile
device or devices associated with a user. For example, a heart rate
sensor (or other type of sensor) may be provided in a glove, or
attached to a wristband, chest strap or similar.
[0035] The physical exertion data is associated with the movement
data. Thus, the physical exertion data is indicative of the
physical exertion data when performing a given movement, e.g. at a
given time and position as indicated by the movement data. This
enables the relatively difficulty of traversing the segment to be
determined for that user. The physical exertion data may be
associated with the movement data in any suitable manner which
allows it to be linked to the relevant movement data. In
embodiments in which the physical exertion data is obtained from a
separate device to the movement data, the data may be associated
with the movement data, e.g. at a central server.
[0036] The physical exertion data may be indicative of any measure
or measures of a physical activity intensity of the user. Measures
may be based upon any or all of: a heart rate of the user, pulse,
blood oxygen content, Borg Rating of Perceived Exertion, CO.sub.2
blood saturation, VO2 max value, etc. In preferred embodiments the
physical exertion data is indicative of at least a heart rate of
the user. The heart rate might be a maximal heart rate or any other
heart rate value. The physical exertion data is preferably obtained
using a heart rate sensor associated with the user. The heart rate
sensor is, in embodiments, associated with, e.g. operably connected
to, a mobile device that provides the movement data e.g. the
portable personal training device of the user.
[0037] In accordance with the invention, the received probe data
from each user is processed using an ability profile for the user.
The ability profile is an individual profile for the user
indicative of the ability of the user to traverse the one or more
navigable segments. The ability profile may be generic to all
navigable segments of the electronic map or may be dependent upon
the navigable segment. For example, the profile may vary dependent
upon a category of a navigable segment, e.g. whether uphill, likely
to be muddy, etc, taking into account the individual difficulties
of a user in traversing different segment types. The ability
profile may be a function of one or more factors affecting the
ability of the individual to traverse a navigable segment or
segments. The ability profile may be a function of one or more of:
a fitness level of the user, an experience level of the user, an
agility level of the user, a stamina level of the user, physical
characteristics of the user, and an equipment type of the user
where applicable. Preferably the ability profile is a function of
at least a fitness level of a user. Thus, in embodiments, the
ability profile may be an ability profile. While the term "fitness
profile" may be used herein, it will be appreciated that, unless
the context demands otherwise, the term may be replaced by the
broader term "ability profile".
[0038] The ability profile data for individual users is used in
processing the probe data obtained from the users to enable the
cost data for traversing a given navigable segment or segments to
be generated. The use of the ability profile data provides a way of
compensating for individual differences in difficulty in traversing
a segment as indicated by the probe data to enable global cost data
to be determined for a segment based upon the probe data obtained
from different users when traversing the segment. For example, a
user with a higher fitness level, more suitable equipment and/or
greater experience, may be able to traverse a given navigable
segment with lower levels of exertion than might another with lower
fitness levels, lower experience levels and/or less appropriate
equipment. By taking into account an ability profile of the
different users, the probe data from multiple users may be
appropriately scaled to provide comparable data that can then be
used in providing normalized cost data, e.g. via a suitable
averaging process. In other words, the ability profile data enables
the probe data obtained from different users in relation to the
traversal of a given navigable segment to be compared, and thus
used in deriving normalized cost data for the segment.
[0039] The method may comprise using the ability profile to adjust,
e.g. scale, the probe data associated with the movement data for
the or each user for use in generating the cost data. The method
may comprise receiving probe data comprising movement data relating
to the movement along a given navigable segment and associated
physical exertion data for each of a plurality of users, and
processing at least the physical exertion data of the users using
the ability profile for each user. The method may further comprise
using the processed probe data in generating a normalized cost for
traversing the segment. In embodiments the movement probe data may
alternatively or additionally be processed using the ability
profiles of the users. While the physical exertion data will be
directly influenced by the individual ability of a user, the
movement data may also be so influenced. For example, a speed of
travel, rate of climb, etc as indicated by the movement data will
be influenced by the user's ability and may similarly be adjusted,
e.g. scaled, to compensate for such individual differences to
enable its uses in determining normalized cost data.
[0040] The ability profile data may be obtained in any suitable
manner. The ability profile is preferably a predetermined ability
profile. In some embodiments the ability profile data is based at
least partially on data provided by the or each user. For example
the user may be invited to answer a series of questions to
establish an ability profile. Alternatively or additionally the
profile may be based upon sensed data, e.g. relating to the user's
historic performance, etc. The ability profile data is preferably
obtained from the device associated with the user that provides the
movement data, and preferably from a personal portable training
device of the user.
[0041] References to an "ability profile" or "fitness profile"
herein may be replaced by references to data indicative of the
respective profile.
[0042] The method preferably comprises the step of generating the
normalized cost data for one or more navigable segments of the
electronic map. The method comprises using the processed probe
data, i.e. the probe data processed using the ability profile data,
to determine the cost data. The normalized cost data is data that
may be used in generating a route across the electronic map
comprising the segment or segments. The concept of a cost
associated with the traversal of a navigable segment is known, for
example, in the context of navigating road segments. The cost is
indicative of the difficulty of (or energy expended in) traversing
the navigable segment. The present invention allows cost data to be
reliably obtained for navigable segments in relation to their
traversal by users under their own power, in particular where the
segments are off-road segments. The cost data generated herein is
indicative of a difficulty of traversing the given navigable
segment by a user at least partially under their own power. The
cost data may be generated using at least the processed
[0043] The cost data is normalized in that it allows the relative
cost associated with traversing different navigable segments to be
determined, i.e. the costs associated with different segments to be
compared. The normalized cost data for each navigable segment may
be by reference to a given scale of cost. For example, each
navigable segment may be assigned a cost level according to a given
scale, e.g. of 1-10 or similar, i.e. one having a predefined number
of levels. The cost data provides a way of globally comparing the
difficulty associated with traversing different segments.
[0044] The normalized cost data for a navigable segment may be
determined using the processed probe data and ability profile data
in any suitable manner. The cost data may be a function of multiple
factors including attributes associated with the segment and/or
determined based on the processed probe data.
[0045] The method preferably involves using at least physical
exertion data relating to the travel of each of a plurality of
users along the navigable segment that has been processed using the
ability profiles for the respective users to determine the cost
data, and preferably further comprises using the movement data,
preferably wherein the movement data has been processed using the
ability profiles.
[0046] The movement probe data may be used directly in determining
the cost data, or data derived from the movement data may be used
in determining the cost data. In some embodiments in which the
movement data comprises elevation data the method may comprise
using elevation data in generating the cost associated with
traversing a segment. The elevation data may be used to determine
an elevation change that is used in generating a cost, for example.
Alternatively or additionally speed data determined using the
movement probe data may be used. Of course, other data may
additionally be used in determining the cost. The data may comprise
attributes of the navigable segment e.g. roughness, curvature,
segment length etc. Such data may be associated with the segment in
an electronic map, and/or may be determined using the movement
probe data. The data used in determining the cost associated with
traversing a segment may therefore be known data, or data that is
determined using the probe data.
[0047] In preferred embodiments the cost data is determined using a
machine learning process. The machine learning process may use a
neural network. The system of the invention may therefore comprise
a machine learning system, preferably a neural network, for
determining cost data. Of course, other techniques may
alternatively or additionally be used. The method may, in general,
use a multivariate statistical analysis. The method may comprise
using the processed probe data as an input to the machine learning
process.
[0048] The method may comprise training, and optionally creating,
an estimator model for creating the cost data for use in generating
a route across an electronic map of navigable segments based upon
the processed probe data for the plurality of users. As known in
the art, the estimator model may be trained using data which is
known. The trained model may then be applied to determining new
data. For example, in order to train the model, input data (i.e.
probe data, including movement data and physical exertion data, in
respect of the traversal of a given navigable segment by a
plurality of users, an agility profile for each of the users, known
or desired cost data for the segment, and optionally data about the
surface of the segment, weather data, etc) may be provided to the
model. The input data may provided in an appropriate format for the
model. By repeating this process for multiple navigable segments,
the model may be trained to appropriately map the input data to the
desired output cost data. The model may derive one or more
parameters indicative of a relationship between the input data and
the cost data. Thus, for example, one or parameters are derived
such that when (determined) cost data for a segment is applied with
a user's ability profile, then the time for traversing the segment
and/or the energy (e.g. caloric) expenditure for traversing the
segment matches that as provided in the probe data received from
the user.
[0049] Once the model has been trained, new input data i.e.
relating to a navigable segment for which cost data is not known,
may be input to the model for determining new cost data using the
one or more derived parameters. The model may be re-trained or
updated as needed.
[0050] In accordance with a further aspect the invention provides a
method of training a model for creating cost data for use in
generating a route across an electronic map of navigable segments,
e.g. as described above. The invention in this further aspect may
include any or all of the features described in relation to the
earlier embodiments.
[0051] The method in accordance with any of its aspects may further
comprise the step of storing the generated normalized cost data for
the or each navigable segment. The data may be stored in
association with electronic map data indicative of the segment.
[0052] It will be appreciated that the difficulty of traversing a
given segment may depend upon weather conditions or a time of year.
Cost data for a given segment may be determined that is time or
weather dependent using probe data collected in respect of the
relevant time period or weather condition. The method preferably
comprises the step of generating time and/or weather dependent
normalized cost data for the one or more navigable segments of the
electronic map. The relevant cost data for the time and/or
conditions at which a route is to be travelled may then be used in
generating a route using the cost data. In embodiments the probe
data that is used in accordance with the invention is data relating
to the movement of each user for a given time period and/or set of
weather conditions. The method may comprise using the data to
determine a time and/or weather dependent cost for traversing a
navigable segment. Cost data may be determined in respect of the
same navigable segment for a plurality of different times and/or
sets of weather conditions, e.g. wet or dry weather, winter or
summer, etc.
[0053] Thus, in embodiments, the method further comprises obtaining
at least one of: surface type for at least a portion of the path
traversed by a user (and for which probe data is obtained);
historic (e.g. the past week, day, hours, etc) and current weather
conditions for the geographic area traversed by a user (and for
which probe data is obtained).
[0054] The generated normalized cost data may be used in various
manners. In some embodiments the method further comprises using the
normalized cost data to provide a suggested route to a user.
Preferably the suggested route is based on an ability profile of
the user and one or more user specified parameters. The user
specified parameters may include one or more of: activity time,
distance, start position (or "origin"), end position (or
"destination"), segment (i.e. surface) type, and level of physical
exertion. An appropriate level of difficulty of a route that can be
traversed by a user can be established using the ability profile of
the user and one or more of the user specified parameters, with the
cost data then being used to determine a route meeting the
difficulty criteria and any required user specified parameters. The
route may comprise a set of one or more navigable segments of the
electronic map for which cost data has been determined. The route
is preferably a trail, i.e. an off-road route. The method may
further comprise providing a set of navigation instructions guiding
a user along the suggested route.
[0055] In embodiments in which the time and/or weather dependent
cost data is determined, the method may comprise providing a
suggested route to a user using cost data in respect of a time
period and/or weather condition corresponding to a current time or
weather condition, or a time or expected weather condition
applicable at the time when the route is to be travelled. In some
embodiments in which the current weather condition or expected
weather condition is not known, weather dependent cost data may be
selected for use in determining the route based on historical
weather conditions. For example, appropriate cost data for weather
conditions expected based on historic weather conditions for the
given time, e.g. of day, month and/or year may be used.
[0056] In accordance with the invention in any of its aspects or
embodiments, preferably the steps of receiving the probe data and
determining the normalized routing costs are carried out by a
central server. The system of the present invention may comprise a
central server arranged to perform any or all of the steps referred
to herein. The received data may be received directly or indirectly
from a device.
[0057] As mentioned above, references to a "trail" herein should be
understood to refer to a path comprising one or more off-road
navigable segments.
[0058] According to a further aspect of the invention, there is
provided a method for determining a route to take along a trail
system comprising:
[0059] acquiring input parameters from at least one participant for
a trail activity, the input parameters comprising: a starting
location and ending location; a type of trail activity; selection
of a routable trail network database which supports the type of
trail activity to be performed; a participant profile of at least
one participant for the type of trail activity to be performed; at
least one of a requested traversal time of the route and the
requested energy consumption to traverse the route; and a maximum
level of exertion to traverse the route;
[0060] determining routing costs associated with potential route
segments as a function of the input parameters, the participant
profile and the at least one of traversal time, energy consumption
and maximum level of exertion, and the routable trail network
database; and
[0061] calculating one or more optimal routes, when feasible, made
up of connected route segments from the selected trail network,
said optimal routes that are nearest to complying with at least one
of the selected requested traversal time, requested energy
consumption and maximum level of exertion.
[0062] The participant profile may include indexes for fitness,
agility and stamina for a given trail activity.
[0063] The at least one participant may comprise two or more
co-participants and where the calculated one or more optimal routes
is calculated based on at least one of a greatest traversal time, a
greatest energy consumption and greatest level of exertion among
the participants for the trail activity to be performed who are
physically capable of traversing the route.
[0064] A potential route may not be considered should the
calculated energy consumption for the participant intending to
traverse the potential route exceeds a maximum value for the
participant.
[0065] According to another aspect of the invention, there is
provided a method to develop a trail routing map database for a
plurality of trail usages and specific to the abilities of a
participant or group of participants comprising:
[0066] identifying one or more trail usages;
[0067] collecting sensor data associated with the traversal of
trail segments for a given trail usage and trail system;
[0068] collecting baseline participant entered input;
[0069] collecting participant entered data during trail
traversal;
[0070] mapping the trail system in 3D dimensions based on sensors
associated with trail segment geography and characteristics;
[0071] developing routing cost model associated with trail segments
based on at least one of: the sensor inputs during trail traversal;
direct participant entered base line and data entered during trail
traversal; and the map of the trail segments traversed.
[0072] The trail usages may comprise one or more of: running,
hiking, boating, biking, skiing, snowshoeing, bridle, snowmobile,
and motorcycle.
[0073] The sensor data may comprise at least one of the following:
trail segment geography and characteristics; environmental
conditions in the vicinity of the trail system; and participant's
bodily functions.
[0074] The trail segment geography and characteristics may include
at least one of latitude, longitude and elevation at intervals
along the segment.
[0075] The environmental conditions may comprise at least one of
real-time weather and historical weather information.
[0076] The participant's bodily functions while traversing a
segment may include at least one of heart rate, pulse oximetry,
body temperature, and acceleration (pref. in 3-component) at
intervals along a segment.
[0077] The baseline participant entered input is a measure of at
least one of the following: the participant's physical
characteristics; and the characteristics of equipment used.
[0078] The participant data entered during trail traversal may
include at least one of: hydration information, caloric intake, and
equipment status.
[0079] The participants physical characteristics may include at
least one of: gender, age, resting and maximal heart rate, V02 Max,
blood oxygen content, Borg Rating of Perceived Exertion, and
CO.sub.2 blood saturation.
[0080] The developing of the routing cost model may be performed
using multivariate statistical analysis and/or machine learning
techniques.
[0081] Principal component analysis may be performed on the routing
cost model such that inputs which have little or no effect on the
outcome of the model are eliminated from the model, and the model
is subsequently recalculated.
[0082] The routing costs may be associated with indexes of
participants fitness, stamina, and agility and a participant's
physical characteristics.
[0083] According to another aspect of the invention, there is
provided a method of establishing a fitness profile of a
participant for a given trail activity comprising:
[0084] collecting sensor input which measure bodily functions
during varying levels of exertion over known trails segments with
known geometry and known routing costs associated with the sensor
types being collected;
[0085] collecting direct participant entered input with respect to
at least one of the following: the participant's physical
characteristics, the presumed level of exertion during trail
traversal, energy consumption during the trail traversal, and the
characteristics of equipment used;
[0086] creating a fitness profile using machine learning techniques
to develop a functional relationship between the sensor inputs and
participant entered input and the routing costs of the routes
taken.
[0087] According to another aspect of the invention, there is
provided a routable trail map database used in a navigation device
comprising: trail segment geometry, elevation, associated routing
costs, and metadata; wherein routing costs for a given trail
activity incorporate: an indication of user's agility, fitness and
stamina for a given trail usage; and an indication of exertion in
the form of energy consumption that is required to traverse a trail
segment in a given direction.
[0088] The routing costs for a given trail activity may be
expressed as a polynomial function used to predict at least one of
participant traversal time and energy expenditure, wherein the
variables of the polynomial function include at least one of:
parameters indicative of the trail segment geography; environmental
factors; indexes indicative of participant, agility and stamina for
a given trail usage, and wherein the multiplicative constants
applied to the variables are weighting functions determined based
on multivariate analysis of sensor measurement by a plurality of
participants traversing the trail segments.
[0089] The at least one sensor measurement may be taken from the
list of: heart rate monitor; pulse oximeter; resting heart rate;
and oxygen content of the blood.
[0090] The parameters indicative of trail geography include at
least one of: average slope; maximum slope; elevation gain; average
curvature; roughness; elevation loss; trail length; trail length
up; and trail length down.
[0091] According to another aspect of the invention, there is
provided a method of maintaining a trail map database and
associated routing costs wherein metadata associated with the
routing costs and geography of the trail system comprises: the
currency of the input data used to create the geography and routing
costs; and the validity of the predictions for travel time and
energy consumption for trail traversal, wherein the geometry or
routing costs are updated when the quality falls below a threshold
by:
[0092] incorporating new sensor input and indices into the
determination of the routing costs and trail segment geography;
[0093] removing data that is older than a selected threshold age
from the determination of the routing cost and trail segment
geography;
[0094] updating the geometry of maps when a plurality of
participants deviate from the stored geometry for a given trail
segment;
[0095] utilizing principal component analysis to remove sensor
input to the routing costs functions that has minimal impact;
and
[0096] adding new sensor input to the routing costs which
positively impact the quality of prediction.
[0097] According to another aspect of the invention, there is
provided a method of determining the influence of equipment or
technique during trail activity comprising:
[0098] developing a personal fitness profile for a single
participant or small group of participants;
[0099] developing trail routing costs for trail segments using the
fitness profile for the single participant;
[0100] incorporate new input on techniques or equipment into the
multivariate analysis; and
[0101] identify the influence that the new input exerts on the
analysis.
[0102] According to another aspect of the invention, there is
provided a method of providing route directions to a participant of
a trail activity using a navigation device comprising:
[0103] determination of a route using the above described
method;
[0104] starting the route;
[0105] monitoring the elapsed time along the route and the
participant's vital signs;
[0106] apprising the participant using one of voice commands or
graphic display during trail traversal of at least one of: turn
directions; level of exertion; deviation from the at least one of
travel time predicted and energy consumption predicted during the
route; warning if the stamina level will be exceeded if the route
is completed; reminder to hydrate or consume calories; elapsed
time; and time to finish.
[0107] According to an aspect of the invention, there is provided a
trail navigation and routing apparatus comprising:
[0108] a processor loaded with software to perform any of the above
described methods;
[0109] a memory containing a routable map database;
[0110] sensors that can measure at least one of location,
elevation, heart rate, pulse oximetry, body temperature,
acceleration, and speed and that can communicate those readings to
the processor; and
[0111] an output device to communicate directions and statistics to
a users of the apparatus.
[0112] In some aspects or embodiments at least, the present
invention may provide a method and apparatus to create, and
maintain a digital trail map that is routable that would enable a
participant to determine a route to take for a trail system unknown
to the participant for a selected type of activity where the
selected route would be traversable for a given energy expenditure
at a certain level of effort and/or in a given amount of time or
where an accurate traversal time and/or energy expenditure could be
estimated going from a starting point to a destination along the
trail system.
[0113] Examples of trail usage are, among other things, recreation
or commuting and sports training. Trail systems are typically used
by pedestrian, hikers or runners or for non-motorized vehicles such
as bicycles, or skis and motorized vehicles such as ATVs
(all-terrain-vehicles), snowmobiles and dirt motorcycles. Water
routes including canoeing and kayaking routes are also included.
This class of transportation system has unique routing costs
associated with it that not only involves aspects of the trail
system, but also physical traits of the user of the trail system
and aspects of the equipment used on the trail and also weather.
Weather is especially important for non-motorized activity as it
has a significant impact on travel-time, such as headwind when
running on a road.
[0114] In some aspects or embodiments at least the invention may
provide the ability to track a participant's fitness and agility
for given trail activities and include these parameters as part of
the routing cost used to determine a route to take. Determining and
updating routing costs including: energy expenditure to navigate a
trail segment, effect of environmental factors, such as weather, on
that energy expenditure, and participant personal fitness and
agility, are determined and updated using multivariate or machine
learning analysis.
[0115] In some aspects or embodiments at least the invention may
involve using a variety of sensor information such as
accelerometers, GPS probe traces, heart-rate monitors, pulse
oximeters and others and evaluating each input's statistical
correlation and relevance and weighting factors for their usage in
updating routing costs.
[0116] In some aspects or embodiments at least the invention may
provide an apparatus to allow a participant to plan a trail route;
to assist in trail mapping and ranking; and to be used as a
training device and to provide trail guidance is part of this
invention. The apparatus is further configured with sensors such as
a GPS, heart-rate monitor and pulse oximeter. The apparatus can log
these inputs and upload to a central server.
[0117] In some aspects or embodiments at least the invention may
assist in competition between a participant and a virtual
competitor or another actual competitor and to handicap competitors
based on agility and fitness profiles
[0118] In some aspects or embodiments at least the invention may
provide a uniform system to develop routing cost data for many
trail uses and for different trail systems and for different
individuals. This enables individual users to be able to use a
different trail system that they previous have no experience with
and pick a route for recreation, commuting or sports training that
will provide them with a realistic travel time or caloric
expenditure and a selection of routing that does not exceed the
participants technical ability or desired level of effort. To come
up with a routable trail map, aspects of probe processing, crowd
sourcing, social networking and geographic information systems come
into play. Because many trail systems change with great frequency:
new trails added, trails are rerouted, trails degrade over time and
improvements are made to the trail, the map and the routing costs
are always changing, the mapping/routing system has to be dynamic.
Likewise an individual's fitness level, any injuries an individual
may be carrying, and an individual's technical skill level are all
dynamic as well.
[0119] In some aspects or embodiments at least the invention may
provide a system to evaluate the effect of equipment on performance
of a competitor.
[0120] In some aspects or embodiments at least the invention may
provide route selection further based on collective profiles of a
group of participants that are using the same trail
simultaneously.
[0121] In some aspects or embodiments at least the invention may
allow track historic routes taken by participants to be tracked and
to incorporate this knowledge into new suggested routes for that
participant so as to not duplicate former routes.
[0122] In some aspects or embodiments at least, a system and
apparatus is presented that is designed to map, route and maintain
a routable digital map database for trails usage (land or water):
the trails being for fitness or commuting and generally a human
propelled vehicle or simply a pedestrian hiker or walker. The
apparatus comprises a distributed network of personal navigation
devices connected periodically to a central processing facility
which collates and redistributes information. Routing cost
(traversal time and/or participant exertion) comprises weather and
human factors. As this invention can be used for human powered
vehicles or pedestrians, the cost of performing a route is highly
tied to the ability of the pedestrian or operator of the human
power vehicle. In order to adequately determine the time and/or
energy output to traverse a trail (or other routing cost) for a
trail system, it is necessary to be able to rank participants
ability when compared to others.
[0123] Advantages of these embodiments are set out hereafter, and
further details and features of each of these embodiments are
defined in the accompanying dependent claims and elsewhere in the
following detailed description.
BRIEF DESCRIPTION OF THE DRAWINGS
[0124] Various aspects of the teachings of the present invention,
and arrangements embodying those teachings, will hereafter be
described by way of illustrative example with reference to the
accompanying drawings, in which:
[0125] FIG. 1, consisting of FIGS. 1A and 1B, shows how an
embodiment where artificial neural network model is used to develop
routing costs for trail routing;
[0126] FIG. 2 is a schematic illustration of electronic components
arranged to provide a portable personal training device;
[0127] FIG. 3 shows an embodiments of the device of FIG. 2, wherein
the device is in the form of a sports watch;
[0128] FIG. 4 describes the process when the effect of addition
parameters is to be evaluated; and
[0129] FIG. 5 illustrates the routing process when a participant
wishes to traverse a trail.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0130] As described above, in its preferred embodiments at least,
the present invention relates to determining costs associated with
navigable segments that are off-road segments, and which costs may
be used in generating routes through a network of such segments.
For example, the routes may be in the form of trails through the
network of off-road navigable segments. A trail as used herein
refers to a path comprising one or more off-road navigable
segments. The segments making up the trail may be referred to as
"trail segments". While the preferred embodiments will be
illustrated by reference to trails and off-road navigable segments,
it will be appreciated that the invention is applicable to other
types of navigable segments and paths through a network of
navigable segments, which may or may not be off-road segments.
[0131] Some fundamental differences of routing for trails, i.e.
paths comprising off-road navigable segments as opposed to paths
through a vehicular road network are: aspects of the participant
fitness and skill have to be taken into account; equipment may
matter substantially; environmental conditions such as weather
matter more and the direction of travel matters due to the effects
of gravity and friction and weather.
[0132] In embodiments it is desired to estimate the travel time to
traverse a trail segment or group of trail segments for a given
participant. The traversal time will be a function of an individual
participant's physical condition, stamina, level of effort, and
previous activity prior to traversal.
[0133] Agility, or how skilled a participant is at performing a
particular trail activity, will influence the amount of energy
required to traverse a trail as well as the mass of the participant
and the fitness of the participant. A less skilled participant may
require addition energy output, or may not be able to traverse a
trail that requires more skill than the individual has. Energy
output required by a participant to traverse a trail is primarily a
function of gravity and friction, but is also influenced by the
agility of the participant. However, trails are not constant and
the weather will greatly influence friction. The stamina of a
participant is also important. If a trail segment is traversed
towards the end of an arduous workout, the participant's agility
could be less and the level of stamina may go down.
[0134] If you know: the energy required to traverse a trail; the
fitness and agility of a participant; how the weather influences
this energy requirement to traverse the trail; and the weather
effect on the ability of the participant, then the time required to
traverse a trail and the energy output needed can be predicted.
Finally, a participant may not want to maximize an effort to
traverse a trail, so not only the type of trail activity must be
considered, but the level of effort, for example recreational,
training, or racing which could be defined, for example, as a
percent of maximum heart rate.
[0135] It is desired is to derive and maintain a routable trail
database and enable use of that database for routing and sports
training and equipment evaluation. The factors needed to predict
traversal time and caloric burn, based on predetermined route costs
for a given activity, can be broken down as follows: [0136]
Participant Ability: Level of Effort (recreation, training,
racing); Fitness; Strength; Agility; Endurance (or Stamina);
Modifiers--Weather (real-time, historical), Equipment, Hydration,
Caloric input [0137] Trail segment degree of difficulty: Energy
expenditure; Agility
[0138] Agility is used herein to mean a participant's skill at the
particular activity. If you are an expert at a particular sport,
you may be able to use less energy to traverse a trail, than
another participant that is equally fit but less skilled. Likewise,
if you have better equipment that allows you to perform better,
then this would affect an agility rating.
[0139] Initial values and changes to the map and routing costs can
be "learned" by applying multivariate analysis to observations of
sensor measurements of the trail system and individual traits and
the weather.
[0140] Due to the different trail uses, the significant amount of
trail systems that exist, and the lack of any commerce occurring on
most trail systems, the mapping of the trail systems and
determination of routing costs will usually be crowd sourced
meaning that sensor data will be uploaded to a central repository
and a map and routing costs derived either automatically and/or
manually from the probe traces. This has been enabled by the common
usage of personal navigation system and smartphone applications and
a shown willingness of sporting participants to upload information
and compare statistics with one another. The present invention in
its embodiments involving the determination of a cost associated
with navigable segments may or may not involve the initial mapping
of the network of segments. In some arrangements, the navigable
segments, e.g. trail segments, of the system may already be mapped,
and the method may use a pre-existing electronic map of trail
segments.
[0141] There is also an aspect of directionality for routing cost
on a trail system--gravity and friction play a significant factor
for human powered activities and also for machine performance on a
trail system.
[0142] In order to accurately predict the transit time for a route,
it is vital to have a uniform rating system for the degree of
difficulty for individual trail segments and the fitness, endurance
and agility of the participant. These costs cannot be measured
directly in most cases and therefore, they must be inferred from
input either by direct observation and assessment (tends to be
subjective) and sensor measurement. Consequently the rating system
must be defined universally and be able to cover the extremes of
fitness and difficulty. An example of a ranking system for an
activity is the rating system for white water kayaking where
sections of the river are rated from I to V where class V implies
that to fall out of the boat and being forced to swim is
potentially life threatening. This system although highly useful,
is limited due to its empirical nature. Not to mention, the class
will change based on how much water is coming down a particular
portion of the river. Likewise this system does not have enough
precision if things like equipment evaluation are to be performed.
There would have to have an order of magnitude more precision such
as class 1-50 instead of class I-V. The rating system for a trail
may ideally span "any one can do it" to "no one can do it". The
rating system for a person may ideally span no effort to impossible
effort; no agility to unobtainable agility.
[0143] Because sensors that are currently available are attached to
a participant or at least to the equipment of the participant, you
cannot divorce the observer from the experiment, meaning if you are
trying to measure the degree of difficulty for a trail segment, the
fitness and agility level of the participant will affect the
measurements. In addition the endurance of a participant is also
important. Fitness level may go down during an extended outing if
endurance is low.
[0144] Trail specific (energy expenditure (or fitness) and agility)
routing costs can all be tied to the influence of gravity and
friction. One could measure friction indirectly, for example, in
cycling, you could measure torque using strain gauges attached to
wheel axle, however, slippage of the wheel is not only due to the
slipperiness of the surface, but how the participant centres
his/her weight over the rear axle. Examples of observations that
can be used to determine a degree of difficulty are slope and the
length of that slope. Another example would be simply record net
elevation gain for a particular segment. An example of a
measurement to correlate with degree of difficulty in terms of a
user's agility or expertise in a particular trail activity would be
changes in acceleration as monitored by accelerometers in a
smartphone or other device warn by a participant. If acceleration
is constantly changing in direction and orientation, this could be
an indication of a bumpy surface of the trail.
[0145] The key factor here is that the user cannot be separated
from the trail when making measurement and in general a degree of
difficulty or a fitness and agility level cannot be measured
directly or do not correlate well to a single type of
measurement.
Multivariate Analysis and Machine Learning
[0146] In embodiments, a model of a complex system comprising
trails, participants, weather and activity is created. As no two
activities are the same, no two participants are the same, and the
interactions of any two components are never the same, it is
impossible to derive routing costs directly--few of the routing
costs can be directly measured; and due to the dynamic nature of
the process, these routing costs will constantly be changing. In
addition, the factors (parameters being measured) which have a
greater correlation with routing costs may be different depending
on the type of trail usage. Also as the system is complex, a form
of multivariate analysis may find better correlations than can be
had with conventional statistical means or direct observation.
[0147] In the context of this application, multivariate analysis
consists of any analysis that takes a variety of measurements or
observations related to an activity and of the participants of that
activity and develops a model combining the influence of these
measurements or observations to predict an outcome. In this case
the outcome to predict is the transit time for a given participant
on a given route and/or the energy consumption to traverse the same
route.
[0148] There are a variety of statistical means to perform
multivariate analysis--analysis of several variables used in
combination to predict an outcome. In an embodiment, an outcome is
to relate the physical characteristic of the trail segment,
environmental parameters and fitness and agility characteristics of
several participants to the time to traverse the trail segment and
the number of calories needed to traverse the segment for a given
participant for a given activity.
[0149] Regardless of the statistical means used, the procedure for
developing a rating system are similar and a person skilled in the
art would be able to use a variety of analysis algorithms based on
this narrative. In an embodiment as shown in FIG. 1, how an initial
model is developed is described for a back propagation artificial
neural network. The objective of the analysis is to predict for a
given trail segment and given activity and given participant:
travel time, energy consumption, maximum level of exertion.
[0150] The method comprises the following steps:
1. Collect Data
[0151] Measure fundamental quantitative and qualitative indicators
of fitness for several participants of the same activity such as
minimal and maximal heart rate, pulse oximeter readings, body mass
index, height, weight, weight of equipment, and age--step 100.
[0152] Collect data for the above participants during traversal of
several trail segments of varying degrees of difficulty--step 102.
Examples of sensors used to collect information include GPS for
geographic location and elevation, heart-rate sensors, 3-axis
acceleration and pulse oximetry for a measure of oxygen content of
the blood--all at intervals over time. [0153] Collect weather and
other environmental data such as precipitation, precipitation for
the previous week, wind speed/direction, temperature ground cover
(leaves, snow, etc) and any similar measurements that may influence
the time to traverse a trail segment or the energy consumed. [0154]
Create several variations of averaged or integrated sensor
measurements over the length of each segment that can be input into
an artificial neural network (or similar multivariate
technique)--step 104. Examples are average speed, average
acceleration, peak speed, peak acceleration. Average acceleration
parallel to the earth surface, both in the direction of travel and
perpendicular to the direction of travel and perpendicular to the
surface; net elevation gain, total elevation, elevation loss,
maximum slope, average slope, average altitude; starting
heart-rate, average heart rate, minimum heart rate, total
heartbeats per trail segment, average blood oxygen level, maximum
blood oxygen level. [0155] Alternatively, general qualitative
indexes can be created (not shown) for input into the neural
network, for example the amount of calories expended to traverse a
trail segment could be approximated by an index based solely on
heart rate over time or solely on pulse oximetry or a combination
of the two. This would allow multiple sensor types to be used to
derive the same index which in turn could all be used in the method
simulation as an input variable instead of the individual sensor
measurements. Of course the issue is that not all sensor types are
as good at indirectly measuring a specific index so this would have
to be accounted for with metadata to indicate the precision the
measurements used to create an index.
2. Create an Artificial Neural Network
[0155] [0156] In this embodiment, to build a back propagation
neural network, training and validation datasets must be assembled
from the above collected data or derivations of the data or
indexes. Several combinations of inputs need to be implemented in
the neural network and the quality of the resulting predictions for
each configuration needs to be assessed. Care needs to be taken to
limit redundant derived indexes or sensor inputs that have similar
effects on the neural network outcome. For example, it may not be
desirable to use both an average heart rate and an index for
calories burned that is derived from heart rate in the same neural
net. In terms of principal component analysis, the measurements or
indexes should be orthogonal--meaning they should represent
mutually exclusive parameters when possible. [0157] Each network is
trained to predict the desired output: traversal time and caloric
expenditure for each segment.
3. Training and Validating the Network
[0157] [0158] A subset of inputs is selected as a training set.
Initially weights are selected for each neural connection or random
weights are input depending on the particular algorithm in use. The
network is then iterated in a learning mode and the weights
associated with each input are adjusted to optimize to prediction
of the output. Next, the optimized neural net is used with the
validation set and the variance between the predicted output and
the measured output is quantified and a model with the minimum
error of prediction is selected. Alternatively a sensor suite that
has a slightly less good prediction, but yet is less expensive or
more reliable to operate could be selected--steps 106-112. [0159] A
principal component analysis is performed and if any input
parameters have very little influence, they can be discarded and a
new neural net constructed and trained without using the discarded
parameters--step 114. [0160] Error functions are created such that
if all sensors of import were not available when a particular trail
segment was characterized, an estimate of the quality of prediction
for that segment can be made based on the relative influence of
each parameter that is being used--step 116. [0161] Use the model
for traversal time prediction and, if desired, display a measure of
the quality of the prediction.
4. Determine the Level of Difficulty
[0161] [0162] Not all participants will be able to navigate every
trail segment--it may be too difficult for their level of fitness
or skill. Once a relationship for travel time and energy
consumption is developed, it can be refined by developing a
relationship to define when a trail exceeds a participants
abilities. Examples of relationships to be observed is if
participants never traverse a trail; if there are large
accelerations in directions other than the direction of travel
based on accelerometer reading or similar; if they fall a lot; and
if there are frequent stops during traversal. A degree of
difficulty can then be assigned to trail segments which can then be
used as an addition routing cost parameter--step 118.
5. Use the Network
[0162] [0163] Once the initial neural network is created and
validated, add additional trail segments and participants. When new
segments or participants are added, acquire sensors measurements
which then are added to the database.
6. Update the Network
[0163] [0164] Periodically perform back propagation or similar
technique to adjust the weighting factors of the network and to
rank new participants and trail segments. [0165] Periodically
remove older data from the database which is incorporated into the
network calculations. This serves to eliminate changes in the trail
system from affecting routing costs and serves to update a user
profile when fitness levels change. [0166] When new sensor types
come on line, record statistically sufficient data for a trail set
and recalculate a neural network using the new sensor type. Update
the network with the new sensor type if justified by better
prediction. Remove older sensor types that analysis shows that no
longer are relevant (have little influence on predicted travel time
or energy consumption or degree of difficulty). [0167] Given a
quantity and variety of measurements that may or may not be related
to an outcome (a routing cost for example), you can develop
weighting factors for each independent variable to determine a
combination of measurement that provide the best prediction and a
confidence level of how good that prediction is. As more
measurements are obtained, the algorithm learns how to make better
predictions. As other types of measurements are obtained that are
analyzed as to whether they improve the predictions of the model
and if so, are incorporated into the model.
[0168] The goal of this multivariate analysis is to have several
measurements (independent variables) which can be used to predict
an outcome (a routing cost) and ultimately a travel time and energy
consumption. The prediction takes the form of a polynomial equation
where individual parameters are normalized and are multiplied by
weighting factors and summed with the result being the predicted
outcomes. So for a given route and participant, level of effort,
weather and equipment, a time to traverse the trail, the amount of
energy and the agility required to traverse the trail can be
predicted. Next the prediction is compared with the actual travel
time. Back propagation can then occur where the weighting factors
are adjusted (as a function of how much confidence there was in the
weighting factor to begin with) such that the polynomial now better
predicts travel time. Confidence in the weighting factors is
recorded and is a function of the precision and accuracy of the
measurement and how many observations went into the determination
of the weighting factor of the measurement.
[0169] It is understood that not all participant will have the same
probe sensors available, nor may their fitness or agility be
documented as well as other participants, nor may they want to
share this data with the network. A participant could use the
system for routing with no sensors by entering estimates of
required input in a quantitative manner. For example a participant
could be prompted for what they think their fitness level is and
also enter weight and age and the system could infer a heart rate
minimum and maximum.
[0170] In an embodiment a way to optimize prediction is to have
multiple neural networks for different classes of participants for
users of different suites of sensing devices. Then the network is
optimized for the available sensor inputs for the given group.
[0171] In another embodiment, if a participant has fewer
measurements to work with, say for example that no weather data is
available and the participant simply has a heart-rate monitor and a
GPS, once the participant has traversed a statistically significant
number of trails that are well characterized, and a fitness level
is established based on the time it takes to traverse several
different trails, then a future travel-time can be predicted albeit
with less accuracy that a full sensor/measurement suite. The
accuracy, in an embodiment, can be characterized by the statistical
relevance of the sensor readings available in the predictive model
being used.
Map Construction
[0172] The methods of the present invention may utilize a
pre-existing electronic map of navigable segments, or may involve a
step of generating an electronic map.
[0173] The routing on a trail system must start with a routable map
and attribution. The map is considered dynamic or constantly
changing. Where it is conducted, initial map generation can be
derived from one or more probe traces from personal navigation
devices or smart phones or other location sensor device. The trace
(a series of locations and elevation measurements and the
associated time of measurement) is uploaded along with user
observations concerning naming, type of feature, etc. and a
navigable map is constructed using conventional means. In addition,
other sensor measurements and/or calculated indices are associated
with the trace are also uploaded. The location and elevation data
is applied to a mean probe trace which results in a series of trail
segments and nodes where one or more trail segments
terminate/intersect or where there is a substantial change in
routing costs--for example a steep hill.
[0174] It is known in the art how to build a routable digital map
for vehicular traffic. It is also known to upload probe (GPS)
traces from portable navigation devices to a central website for
the purpose of construction of a digital map. Likewise strictly
crowd sourced street maps are well known. For this application an
initial map and further updates to the map can be performed using
similar techniques to crowd source and build a vehicular routable
road network.
[0175] One addition to the techniques used in constructing a map
for vehicular traffic, however, is to have user assistance when
building/maintaining the trail routing system whereby a user
interface on a personal navigation device interactions with the
user asking simple question and/or allowing a user to attribute
probe traces while acquiring them. For example naming or numbering
trail segments and/or confirming the name of an existing trail
segment being traverse or identifying when a node is crossed. This
information can go a long way to improve conflation of probe traces
into an existing database or building one from scratch.
[0176] It should be noted that the overall database can be
distributed over a variety of devices just so long as those devices
are in communication at least intermittently and periodically. For
example the trail map for one trail system need not reside on the
same server as the trail map for another system. Likewise,
individual user/participant profiles can reside on a personal
navigation device or personal computer and for privacy reasons;
statistics on the user can be uploaded periodically and anonymously
(if desired).
[0177] Nodes in the trail system may also be placed at other than
at the intersection of more than two segments or at a dead-end
segment. Nodes may also be placed where the degree of difficulty or
the fundamental routing cost changes appreciably. For example, if a
portion of trail is extremely steep, then it flattens out, it may
be advisable to place a node at the transition location of the
slope. Likewise, if the surface transitions from payment to crushed
stone, for example, this may be observed with a node.
[0178] Various combinations of manual map making, conflation of map
information and/or information from probe traces can be used to
make an initial map. Once the initial map is made, then the map and
routing costs are continually refined with information from probe
traces uploaded from users in combination with user profiles,
monitoring information from the user and environmental data such as
weather. Geometry from the probe traces is constantly introduced
and compared to the existing database. If geometry varies greatly,
then a person uploading the trace may be asked questions such as:
Is this a new trial? If not, does this trace follow trace X going
through a certain node and intersect trace Y? Therefore from a
combination of raw data and simple question and answers, the trail
network map can be continually updated.
Data Collection
[0179] It can be inferred that certain parameters will influence
routing cost more than others, but artificial neural network
analysis may find combinations of factors that are not obvious.
When first collecting data, all commonly available sensor output
should be recorded. This includes GPS probe traces (including time,
location, elevation), heart-rate, and acceleration (from smartphone
accelerometers).
[0180] In embodiments of the invention, positional data is obtained
from portable personal training device, such as sports watches,
having access to Global Positioning System (GPS) data. Such sports
watches are often worn by athletes to help them during their runs
or workouts, e.g. by monitoring the speed and distance of the user
and providing this information to the user. It will be appreciated,
however, that the device could be arranged to be carried by a user
or connected or "docked" in a known manner to a vehicle such as a
bicycle, kayak, or the like.
[0181] FIG. 2 is an illustrative representation of electronic
components of a personal portable training device 200 according to
a preferred embodiment of the present invention, in block component
format. It should be noted that the block diagram of the device 200
is not inclusive of all components of the navigation device, but is
only representative of many example components.
[0182] The device 200 includes a processor 202 connected to an
input device 212 and a display screen 210, such as an LCD display.
The input device 212 can include one or more buttons or switches
(e.g. as shown in FIG. 3). The device 200 can further include an
output device arranged to provide audible information to a user,
such as alerts that a certain speed has been reached or a certain
distance has been travelled. FIG. 2 further illustrates an
operative connection between the processor 202 and a GPS
antenna/receiver 204. Although the antenna and receiver are
combined schematically for illustration, the antenna and receiver
may be separately located components. The antenna may be a GPS
patch antenna or helical antenna for example.
[0183] The device 200 further includes an accelerometer 206, which
can be a 3-axis accelerometer arranged to detect accelerations of
the user in x, y and z directions. The accelerometer may play a
dual role: firstly as a means for determining a motion state of the
wearer at a particular moment in time, and secondly as a pedometer
for use when/if there is a loss of GPS reception. Although the
accelerometer is shown to be located within the device, the
accelerometer may also be a external sensor worn or carried by the
user, and which transmits data to the device 200 via the
transmitter/receiver 208.
[0184] The device may also receive data from other sensors, such as
a footpad sensor 222 or a heart rate sensor 226. The footpad sensor
may, for example, be a piezoelectric accelerometer that is located
in or on the sole of the users shoe. Each external sensor is
provided with a transmitter and receiver, 224 and 228 respectively,
which can be used to send or receiver data to the device 200 via
the transmitter/receiver 208.
[0185] The processor 202 is operatively coupled to a memory 220.
The memory resource 220 may comprise, for example, a volatile
memory, such as a Random Access Memory (RAM), and/or a non-volatile
memory, for example a digital memory, such as a flash memory. The
memory resource 220 may be removable. As discussed in more detail
below, the memory resource 220 is also operatively coupled to the
GPS receiver 204, the accelerometer 206 and the
transmitter/receiver 208 for storing data obtained from these
sensors and devices.
[0186] Further, it will be understood by one of ordinary skill in
the art that the electronic components shown in FIG. 2 are powered
by a power source 218 in a conventional manner. The power source
218 may be a rechargeable battery.
[0187] The device 200 further includes an input/output (I/O) device
216, such as a USB connector. The I/O device 216 is operatively
coupled to the processor, and also at least to the memory 220 and
power supply 218. The I/O device 216 is used, for example, to:
update firmware of processor 220, sensors, etc; transfer data
stored on the memory 220 to an external computing resource, such as
a personal computer or a remote server; and recharge the power
supply 218 of the device 200. Data could, in other embodiments,
also be sent or received by the device 200 over the air using any
suitable mobile telecommunication means.
[0188] As will be understood by one of ordinary skill in the art,
different configurations of the components shown in FIG. 2 are
considered to be within the scope of the present application. For
example, the components shown in FIG. 2 may be in communication
with one another via wired and/or wireless connections and the
like.
[0189] FIG. 3 illustrates a preferred embodiment of the device 200,
wherein the device 200 is provided in the form of a watch 300. The
watch 300 has a housing 301 in which is contained the various
electronic components on the device as discussed above. Two buttons
212 are provided on the side of the housing 301 to allow the user
to input data to the device, e.g. to navigation a menu structure
shown on the display 210. Any number of buttons, or other types of
input means, can alternatively be used as desired.
[0190] The watch 300 has a strap 302 for securing the device to a
users wrist. As can be seen the end of the strap 302 has a hinged
cover 304 that can be lifted up, e.g. as shown in FIG. 3A, to
reveal a USB connector 308. The connector can be inserted into any
suitable USB port for power and/or data transfer.
User Fitness and Agility Initial Determination
[0191] The following provides an example of the way in which an
ability, e.g. fitness, profile may be derived for a user.
[0192] In one embodiment it can be assumed for initial user
profiling that fitness can be measured for all activities by simple
aerobic fitness and further limited to heart-rate monitoring which
is a fair predictor of aerobic fitness.
[0193] A basic embodiment of the invention is described here. It is
assumed that a trail system is mapped in 3 dimensions, both
latitude and longitude and elevation. The map is defined by at
least segments between trail intersections and nodes at the
intersections. Base line fitness measurements are made of at least
one participant which include body mass index, resting and maximal
heart rate weight, height, gender, and age. The heart rate and time
and location of the at least one participant is continually
monitored while the trail network is being traversed. With only the
heart rate monitor, the baseline participant physical parameters
and the geometry of the trail system, a routing model is
constructed using the method previously described.
[0194] Aerobic fitness is generally presumed to be best represented
by Maximal Oxygen Uptake (VO2 Max). One indirect method to measure
VO2 Max is by comparing the heart rate at rest to the maximum heart
rate, but direct oxygen uptake measurement may be feasible in the
future. Heart rate (or pulse) is also an imprecise predictor of
physical strength and more particularly, physical strength of
certain muscle groups. Strength may be more important than aerobic
fitness for certain types of trail activities.
[0195] The result is new type of sensor measurements may be
available in the future, in which case a quantity of the these
measurements would need to be taken for a variety of participants
and a variety of trail segments and a new learning session ensued
to see if the new sensor enhances the prediction and/or can replace
other sensors measurements without degradation of prediction
variance. For now, a participant needs to define a resting pulse
and a maximal pulse. These can be measured directly in a stress
test environment or could be estimated via the Karvonen method or
other method known in the art.
[0196] As it is anticipated that serious athletic competitors will
want to evaluate training regiments or equipment, the rating system
will have to be both accurate and precise as possible.
[0197] A measure of arterial oxygen saturation by pulse oximeter is
a sensor of particular interest as this in combination with pulse
rate could be a valuable predictor of participant fitness. It is a
further object of this invention to have a pulse oximeter sensor
embedded in a glove to keep the sensor firmly against a fingertip
with the glove further protecting either a hardwire to a wrist
receiver or a Bluetooth or similar wireless transceiver which
relays pulse oximetry to a recorder. Yet another embodiment is to
have a chest strap harness which contains a regional pulse oximeter
sensor singly or in combination with a heart rate sensor. Other
useful measurements include things such as CO.sub.2 blood
saturation.
[0198] Likewise a user can estimate their fitness/agility level
based on their fitness/agility compared to others. Alternatively
the method could adapt the Borg Rating of Perceived Exertion. The
Borg Rating however is a minimum and maximum exertion level for an
individual and is not relative between participants. Therefore it
would need to be translated to heart rate which for the Borg Rating
is determined. There are various means to estimating the heart rate
for various amounts of exertion and a person's age.
[0199] Training athletes or equipment evaluators singly or in
groups may wish to devise their own multivariate analyses so they
can observe subtle differences in energy expenditure that are a
function of equipment or technique (FIG. 4). So these groups or
individuals may choose to include only trails segment in a study
that are highly characterized with lots of probe traces and
participant records and to build up a database of additional
measurements and observations with respect to these highly measured
trails. An example for biking would be to monitor, type of bike,
tire pressure, tread pattern, shock pressure, seat position, then
perform personal evaluation of the results to determine the effect
on energy efficiency. Likewise a road biker may monitor pedal
cadence, put this in the network and observe the effect on energy
efficiency; if a participant stops and eats at intervals or not; if
a person has drunk a certain amount of water. If the neural
network, prior to incorporating these factors, is a good predictor
of energy consumption, then adding these independent variables to
the network should make it observable whether they make a
difference. As in FIG. 4, for well characterized participants, a
predictive model can be developed for a single participant and
things like hydration, caloric input and equipment can be
monitored--step 400. The statistical significance of the new input
can be monitored and increases or decreases in performance can be
assessed based on the new inputs--steps 402, 404.
[0200] To accurately predict energy expenditure for a given route,
real-time and historic weather need to be included in the
multivariate analysis calculation as they influence friction forces
on the trail and on the participant, but also may influence the
energy expenditure of the participant. Examples of this are snow on
a trail requiring more energy for hikers because they have to lift
their feet higher and because of decreased traction. Another
example is less efficient energy transfer in certain individuals in
high heat and humidity. Sources of this information can be
participant manual input or feeds from wireless weather
services.
[0201] Information that is useful for trail routing comprises the
following sensors or participant input that can provide an
indication of the parameters below:
1. The Map
[0202] Segments (3-dimensional trace of a portion of a trail system
that ends at nodes) [0203] Nodes (decision points where one or more
segments intersect or where there is a significant change in
routing cost (example: change in slope)) [0204] Historical Weather
(if it rained all last week--trails will be muddy and this will
affect transit times) [0205] Real-time Weather (if it is currently
raining or frosty, transit time will be increase due to
slipperiness). [0206] Temperature/humidity: efficiency of the
user/participant will vary as a function of these parameters 2.
Trail Segment Degree of Difficulty/Trail Base Routing Cost (for
each Activity) [0207] Fitness based (Amount of energy required to
traverse a trail) [0208] Agility based--Is the trail rocky or
covered by tree roots and other vegetation? Are you able to bike
ride a trail segment when others have to walk it) [0209] Direction
(forward or backward along a trail segment)
3. User Profile:
[0209] [0210] Fitness level and physical parameters (can be based
on user input initially or sensor input such as heart rate
monitoring or comparison to other user statistics for a given trail
segment and given conditions [0211] Weight, Body mass index, age,
gender [0212] Agility/Skill Level (how is your balance, reaction
time, cadence, etc) [0213] Endurance (how long can you sustain a
certain level of exertion?) [0214] Equipment Profile: (what kind of
tires, gear ratio, running shoes, clothing, etc are you using; what
horse you are riding?) And for individual outings: [0215] Training
mode or recreation; group or individual [0216] Hydration/Calorie
Intake [0217] Environmental Factors (influence both the trail
degree of difficulty and the user profile), such as: wind speed and
direction; temperature and humidity; historical precipitation
(weekly); and current precipitation.
[0218] Equipment utilization, replenishment of water and nutrients,
can be kept private as well as the analysis of any performance
enhancement resulting from a piece of equipment or regiment. In an
embodiment, for general purposes, only an average fitness and
agility level for an individual needs to be uploaded to a central
processor. This information is necessary to constantly improve the
overall quality of rating the degree of difficulty of a trail
segment and rating the fitness level of other individuals.
[0219] A user profile will be able to be normalized among
individual users by establishing a baseline with the usage of
sensors or specific measurement. For example, an initial baseline
fitness measurement could use the President's Council on Fitness,
Sports and Nutrition guidelines for a general indicator, for
example you could measure you change in pulse rate before and after
a brisk mile walk, how many push-ups can you do? Simple measures of
flexibility; measure of body mass index. This will be a starting
point for individuals, but for trained athletes, much more
information will be necessary. For example a sprinter may be able
to do ten times as many push-ups as a distance runner, however both
could be in top physical condition for their sport. With a control
group of individuals, individual trail segments can be rated as to
the degree of difficulty initially by comparing transit times for a
segment (in both directions) for a variety of individual with
differing fitness documented fitness levels.
[0220] Individuals that have a profile backed with a significant
amount of data could become testers of equipment. By comparing the
equipment against a base line without that particular kind of
equipment, improvements in performance could be correlated.
[0221] Gaming and competitions can be set up amongst actual
individuals or an actually individual against a virtual competitor.
Handicaps can be established much as in golf for team competitions.
Training regiments can be created where you select a fitness degree
and the software informs you if you are performing at above or
below the species fitness. Tracking of improvements of fitness can
also be tracked.
[0222] As trail routes are likely to change more often than a road,
the probe traces that go into making a statistically averaged trail
segment need to be aged and older segments should be weighted less
when developing the average probe trace used to represent to trail
segment. How quickly to age a probe will be a function of how many
probe traces there are available. Likewise, probes may be of
variable quality and precision and probe traces of less quality and
precision should be weighted less.
Usage
[0223] Once a trail map and routing cost and personal fitness is
established then the following can be performed as shown in FIG.
5.
[0224] In order to determine a route, a user would input into a
navigation device, for example, the following: length of time
desired for the activity; type of activity; mode of activity
(training, recreation, other); number of members in the group; name
of group members (provided they have a profile); trail system to be
used (this could also automatically be determined based on the
present location or input location); maximum degree of difficulty
(do you want to tax yourself or take at easy)--step 500.
[0225] The routing algorithm would then determine if it is
connected to the network. If it is connected, then the algorithm
looks for updates to the trail system of interest; either base
costs or changes in geometry--then update the local database. The
algorithm will also look for historical and real-time weather
information--step 506. A modifier to the overall base costs is
determined as a function of, for example, rainfall over the last
week (how muddy are the trails), current and predicted rain/snow
for today (how slippery are the trails). Other factors taken into
account by the algorithm include: the weak link in the group; the
size and makeup of the group; a route does not violate the degree
of difficulty; a route should not be the same as one recently
undertaken--steps 502, 504, 510.
Routing
[0226] Given an activity type, trail system, travel time and a
level of exertion (and maximum desired exertion) (typically based
on a percentage of the range between minimal and maximal heart
rate), then a route is suggested.
[0227] While in route, communicate one or more of the following:
[0228] alert when heart rate is higher or lower than desired
(accounting for the energy required to traverse the current trail
segment--may be required to exceed desired heart rate in certain
steep or difficult sections). [0229] give routing directions at
intersections (nodes) (audio or visual) [0230] alert when extra
exertions will be required [0231] display map view of the trail
system--colour coding energy consumption per trail segment or
maximum slope per trail segment.
[0232] Another aspect of this invention is how to provide the
routing directions. It has already been shown in vehicles that a
navigation device can contribute to driver distraction. When
dealing with a sporting activity that requires even more
concentration than driving, distraction is even more of an issue.
In a training scenario, it is not convenient to stop and look at a
map. In addition, since this system would be used primarily
outdoors, being able to view a screen in direct sunlight can
sometimes be difficult. Accordingly, directions should be delivered
preferably either audibly or in a heads up display--to avoid
distraction. In addition, to simple: "left turn ahead" type of
directions, information on the exertion level for the upcoming
segment could be provided. For example, if you are coming to a
steep hill that continually goes up for 3/4 of a mile, this
information could be conveyed to the participant so that they can
pace their activity. If the participant has an extensive profile on
file and a heart rate monitor was being used, the routing algorithm
could predict the pace based on the predicted stamina of the
participant and the duration of the exertion. This system is
designed to work on portable devices equipped with a GPS or other
location sensor, and optionally heart rate monitors or other
sensors that can measure energy output of a participant. The
portable device, for example can communicate with one or more
servers to share information and calculations. The device can be
configured to perform routing calculations locally on the device or
get them from a server. The device needs at least one of a display,
voice communications or text communications. It may also need to be
configured to acquire trail information by manual input in certain
embodiments.
[0233] An embodiment of a method of the invention for determining
costs associated with the traversal of off-road navigable segments,
e.g. trail segments, of a network of such segments represented by
an electronic map will now be described. The electronic map and the
segments thereof may be known, or may be generated in a further
step using probe data relating to the movement of users with
respect to time in a geographic region comprising the segments. The
probe data may be collected from devices associated with users in
the same manner described in relation to determining the cost
data.
[0234] Probe data is collected from personal portable training
devices associated with different users traversing a given
navigable segment of the network. The probe data obtained from each
device includes time stamped position data, including elevation
data. The probe data additionally comprises data indicative of the
exertion of the user e.g. a heart rate of the user associated with
the time stamped position data. The exertion e.g. heart rate data
is obtained from a suitable sensor associated with the training
device, and is indicative of the level of exertion of the user over
time as they traverse the segment.
[0235] The probe data is collected at a server. An ability, e.g.
fitness, profile has been set up for each user. The fitness profile
may have been established based upon a questionnaire to which the
user has responded, or based on historic performance, etc. Before
being used to determine a cost for traversing the given navigable
segment, the probe data for a user is processed using the ability
profile data for the given user to which the probe data relates. In
this way, the probe data may be adjusted to compensate for
variation in, for example, fitness levels between different users
before it is used to determine a cost for traversing the
segment.
[0236] The processed probe data obtained from each of the users
when traversing a given segment is then used to determine a
normalised cost function for traversing the segment. For example,
the segment may be assigned an integer in a given range, e.g. 1 to
5 indicative of the relative difficulty (i.e. exertion involved in
traversing the segment). This may be achieved using a suitably
trained machine learning technique, e.g. neural network.
[0237] A normalized cost may then be determined for other segments
of the electronic map in a similar manner, and the cost data
associated with each segment. The determined cost is associated
with each segment in a similar manner to that known for road
segments to enable the cost to be used in generating a route
through the network of segments.
[0238] The cost data may then be used to provide a suggestion of a
route through the network of segments for an individual user. For
example, the cost data may be used to propose a route that meets a
given level of exertion range specified by the user, or which is
deemed suitable by reference to an ability, e.g. fitness, profile
of the user. The route may alternatively or additionally be one
meeting a time, or distance criteria specified by a user.
[0239] It will be appreciated that the difficulty of traversing a
given segment may depend upon weather conditions or a time of year.
Cost data for a given segment may be determined that is time or
weather dependent using probe data collected in respect of the
relevant period. The relevant cost data for the time or conditions
at which a route is to be travelled may then be used in generating
a route using the cost data.
[0240] It will be appreciated that embodiments of the present
invention can be realised in the form of hardware, software or a
combination of hardware and software. Any such software may be
stored in the form of volatile or non-volatile storage such as, for
example, a storage device like a ROM, whether erasable or
rewritable or not, or in the form of memory such as, for example,
RAM, memory chips, device or integrated circuits or on an optically
or magnetically readable medium such as, for example, a CD, DVD,
magnetic disk or magnetic tape. It will be appreciated that the
storage devices and storage media are embodiments of
machine-readable storage that are suitable for storing a program or
programs that, when executed, implement embodiments of the present
invention. Accordingly, embodiments provide a program comprising
code for implementing a system or method as claimed in any
preceding claim and a machine readable storage storing such a
program. Still further, embodiments of the present invention may be
conveyed electronically via any medium such as a communication
signal carried over a wired or wireless connection and embodiments
suitably encompass the same.
[0241] All of the features disclosed in this specification
(including any accompanying claims, abstract and drawings), and/or
all of the steps of any method or process so disclosed, may be
combined in any combination, except combinations where at least
some of such features and/or steps are mutually exclusive.
[0242] Each feature disclosed in this specification (including any
accompanying claims, abstract and drawings), may be replaced by
alternative features serving the same, equivalent or similar
purpose, unless expressly stated otherwise. Thus, unless expressly
stated otherwise, each feature disclosed is one example only of a
generic series of equivalent or similar features.
[0243] The invention is not restricted to the details of any
foregoing embodiments. The invention extends to any novel one, or
any novel combination, of the features disclosed in this
specification (including any accompanying claims, abstract and
drawings), or to any novel one, or any novel combination, of the
steps of any method or process so disclosed. The claims should not
be construed to cover merely the foregoing embodiments, but also
any embodiments which fall within the scope of the claims.
* * * * *